Purpose: The long-term prognosis of hepatocellular carcinoma (HCC) has improved recently through early detection by periodic surveillance and therapeutic developments. However, clinical trials comparing the options have not been performed, so we sought to estimate the long-term effectiveness of surgical resection or local ablation (LAT) in the initial treatment of solitary small (≤ 3cm) HCC in HCV-related hepatitis and cirrhosis. Methods: We created two Markov models to simulate the clinical course of HCC from first treatment to death from liver disease or other causes. Model A considered complete remission (CR) or non-CR for every HCC health state. In Model B, treatment reduced the transition probabilities from HCC health states without consideration of CR. Survival with HCC depended on maximal size, number and location of tumor nodules. Transitional probabilities were estimated from 381 HCC cases, including 105 small solitary HCC ones reviewed retrospectively. We compared the predicted survival from these two models to observed Kaplan-Meier survival analysis rates for actual cases. Results: As shown in the following table, for patients hypothetically identical to the study population with mean age 65-years, 71% men and 81% cirrhosis, Model A predicted short-term survival rates more consistently similar with those from the Kaplan-Meier survival analysis than Model B. On the contrary, the Model B predicted long-term survival rates more consistently similar to the Kaplan-Meier ones than Model A. Conclusion: The incorporation of complete remission in Markov models affects predicted cancer survival; in this case, long-term survival without CR more closely matched Kaplan-Meier estimates.